HITL Overview
Allow your agent and users to collaborate on complex tasks.
"use client";import React from "react";import { CopilotKit, CopilotChat, useHumanInTheLoop, useConfigureSuggestions,} from "@copilotkit/react-core/v2";import { z } from "zod";import { TimePickerCard, TimeSlot } from "./time-picker-card";const DEFAULT_SLOTS: TimeSlot[] = [ { label: "Tomorrow 10:00 AM", iso: "2026-04-19T10:00:00-07:00" }, { label: "Tomorrow 2:00 PM", iso: "2026-04-19T14:00:00-07:00" }, { label: "Monday 9:00 AM", iso: "2026-04-21T09:00:00-07:00" }, { label: "Monday 3:30 PM", iso: "2026-04-21T15:30:00-07:00" },];export default function HitlInChatDemo() { return ( <CopilotKit runtimeUrl="/api/copilotkit" agent="hitl-in-chat"> <div className="flex justify-center items-center h-screen w-full"> <div className="h-full w-full max-w-4xl"> <Chat /> </div> </div> </CopilotKit> );}function Chat() { useConfigureSuggestions({ suggestions: [ { title: "Book a call with sales", message: "Please book an intro call with the sales team to discuss pricing.", }, { title: "Schedule a 1:1 with Alice", message: "Schedule a 1:1 with Alice next week to review Q2 goals.", }, ], available: "always", }); useHumanInTheLoop({ agentId: "hitl-in-chat", name: "book_call", description: "Ask the user to pick a time slot for a call. The picker UI presents fixed candidate slots; the user's choice is returned to the agent.", parameters: z.object({ topic: z .string() .describe("What the call is about (e.g. 'Intro with sales')"), attendee: z .string() .describe("Who the call is with (e.g. 'Alice from Sales')"), }), render: ({ args, status, respond }: any) => ( <TimePickerCard topic={args?.topic ?? "a call"} attendee={args?.attendee} slots={DEFAULT_SLOTS} status={status} onSubmit={(result) => respond?.(result)} /> ), }); return <CopilotChat agentId="hitl-in-chat" className="h-full rounded-2xl" />;}What is this?#
Human-in-the-loop (HITL) lets an agent pause mid-run to collect input, confirmation, or a choice from the user, then resume with that answer folded back into its reasoning. It's what turns an autonomous workflow into a collaborative one: the agent keeps its context, the user keeps the steering wheel.
When should I use this?#
Use HITL when you need:
- Quality control — a human gate at high-stakes decision points
- Edge cases — graceful fallbacks when the agent's confidence is low
- Expert input — lean on the user for domain knowledge the model lacks
- Reliability — a more robust loop for real-world, production traffic
Two patterns for HITL in CopilotKit#
Install the LangGraph Python SDK
uv add copilotkitpoetry add copilotkitpip install copilotkit --extra-index-url https://copilotkit.gateway.scarf.sh/simple/conda install copilotkit -c copilotkit-channelWire CopilotKit middleware into your graph
For useHumanInTheLoop tool-based HITL, the tool is defined entirely on
the frontend and forwarded to the agent. CopilotKitMiddleware is what
forwards it — drop it into your create_agent call.
from langchain.agents import create_agent
from langchain_openai import ChatOpenAI
from copilotkit import CopilotKitMiddleware
graph = create_agent(
model=ChatOpenAI(model="gpt-5.4"),
tools=[],
middleware=[CopilotKitMiddleware()],
system_prompt="You are a helpful, concise assistant.",
)For the useInterrupt graph-paused pattern, you'll also use LangGraph's
native interrupt(...) primitive inside a graph node — no extra
CopilotKit setup beyond the middleware above.
CopilotKit ships two complementary ways to pause an agent turn and ask the human something. They look similar from the outside (the chat pauses, a custom component appears, the user answers, the run resumes) but they're wired differently on the backend, and each has its own niche.
| Pattern | Who decides to pause? | Backend surface |
|---|---|---|
useHumanInTheLoop | The LLM, by calling a registered client-side tool | A frontend-only tool description (Zod schema + render) |
useInterrupt | The graph, by calling interrupt(...) during a node | A server-side interrupt() call in your LangGraph agent |
Pick useHumanInTheLoop when the pause is an agent-initiated
decision — the model chose to ask the user — and you want the picker UI
inlined into the normal tool-call flow.
Pick useInterrupt when the pause is a graph-enforced checkpoint —
the code path deterministically requires a human answer — and you want
langgraph.interrupt() as the server-side contract.
Pattern 1 — useHumanInTheLoop (tool-based)#
The agent registers a HITL tool on the client with useHumanInTheLoop.
When the LLM calls that tool, CopilotKit routes the call through your
render function, which shows a custom component and calls respond
with the user's answer. The agent sees the answer as the tool result and
continues from there.
import React from "react";import { CopilotKit, CopilotChat, useHumanInTheLoop, useConfigureSuggestions,} from "@copilotkit/react-core/v2";import { z } from "zod";import { TimePickerCard, TimeSlot } from "./time-picker-card";const DEFAULT_SLOTS: TimeSlot[] = [ { label: "Tomorrow 10:00 AM", iso: "2026-04-19T10:00:00-07:00" }, { label: "Tomorrow 2:00 PM", iso: "2026-04-19T14:00:00-07:00" }, { label: "Monday 9:00 AM", iso: "2026-04-21T09:00:00-07:00" }, { label: "Monday 3:30 PM", iso: "2026-04-21T15:30:00-07:00" },];export default function HitlInChatDemo() { return ( <CopilotKit runtimeUrl="/api/copilotkit" agent="hitl-in-chat"> <div className="flex justify-center items-center h-screen w-full"> <div className="h-full w-full max-w-4xl"> <Chat /> </div> </div> </CopilotKit> );}function Chat() { useConfigureSuggestions({ suggestions: [ { title: "Book a call with sales", message: "Please book an intro call with the sales team to discuss pricing.", }, { title: "Schedule a 1:1 with Alice", message: "Schedule a 1:1 with Alice next week to review Q2 goals.", }, ], available: "always", }); useHumanInTheLoop({ agentId: "hitl-in-chat", name: "book_call", description: "Ask the user to pick a time slot for a call. The picker UI presents fixed candidate slots; the user's choice is returned to the agent.", parameters: z.object({ topic: z .string() .describe("What the call is about (e.g. 'Intro with sales')"), attendee: z .string() .describe("Who the call is with (e.g. 'Alice from Sales')"), }), render: ({ args, status, respond }: any) => ( <TimePickerCard topic={args?.topic ?? "a call"} attendee={args?.attendee} slots={DEFAULT_SLOTS} status={status} onSubmit={(result) => respond?.(result)} /> ), });The picker UI is fed a static list of candidate slots — this is just data the demo page owns, so you can swap in real availability, a calendar API, or anything else:
import React from "react";import { CopilotKit, CopilotChat, useHumanInTheLoop, useConfigureSuggestions,} from "@copilotkit/react-core/v2";import { z } from "zod";import { TimePickerCard, TimeSlot } from "./time-picker-card";const DEFAULT_SLOTS: TimeSlot[] = [ { label: "Tomorrow 10:00 AM", iso: "2026-04-19T10:00:00-07:00" }, { label: "Tomorrow 2:00 PM", iso: "2026-04-19T14:00:00-07:00" }, { label: "Monday 9:00 AM", iso: "2026-04-21T09:00:00-07:00" }, { label: "Monday 3:30 PM", iso: "2026-04-21T15:30:00-07:00" },];Pattern 2 — useInterrupt (graph-paused)#
With LangGraph's interrupt() the pause is enforced by the graph
itself: a node calls interrupt({...}), the run suspends, the client
receives the payload, renders a UI, and resumes the run with the user's
answer. CopilotKit's useInterrupt hook is the render contract.
See the useInterrupt deep dive for
the full walkthrough, including the backend tool and render-prop wiring.
"""LangGraph agent for the Human-in-the-Loop (Interrupt-based) booking demo.Defines a backend tool `schedule_meeting(topic, attendee)` that usesLangGraph's `interrupt()` primitive to pause the run and surface astructured booking payload to the frontend. The frontend `useInterrupt`renderer shows a time picker inline in the chat and resolves with`{chosen_time, chosen_label}` or `{cancelled: true}`, which this toolturns into a human-readable result the agent uses to confirm the booking."""from __future__ import annotationsfrom datetime import datetime, time, timedeltafrom typing import Any, List, Optionalfrom zoneinfo import ZoneInfofrom langchain.agents import create_agentfrom langchain_core.tools import toolfrom langchain_openai import ChatOpenAIfrom langgraph.types import interruptfrom copilotkit import CopilotKitMiddlewareSYSTEM_PROMPT = ( "You are a scheduling assistant. Whenever the user asks you to book a " "call / schedule a meeting, you MUST call the `schedule_meeting` tool. " "Pass a short `topic` describing the purpose and `attendee` describing " "who the meeting is with. After the tool returns, confirm briefly " "whether the meeting was scheduled and at what time, or that the user " "cancelled.")# Demo-only fixed timezone. A real app would use the user's calendar +# locale (e.g. zoneinfo.ZoneInfo(user.timezone) and Google Calendar /# Outlook availability); we hardcode Pacific so screenshots are stable._DEMO_TZ = ZoneInfo("America/Los_Angeles")def _candidate_slots() -> List[dict]: """Upcoming candidate slots, relative to "now" so the picker never shows stale dates.""" now = datetime.now(_DEMO_TZ) tomorrow = (now + timedelta(days=1)).date() # Skip a week when the result would collide with `tomorrow` — i.e. # today is Mon (0 days away, picker would show two slots both # labelled "Monday") or Sun (1 day away, picker would show # "Tomorrow" and "Monday" both pointing at the same date). days_to_monday = (7 - now.weekday()) % 7 if days_to_monday <= 1: days_to_monday += 7 next_monday = (now + timedelta(days=days_to_monday)).date() candidates = [ ("Tomorrow 10:00 AM", tomorrow, time(10, 0)), ("Tomorrow 2:00 PM", tomorrow, time(14, 0)), ("Monday 9:00 AM", next_monday, time(9, 0)), ("Monday 3:30 PM", next_monday, time(15, 30)), ] return [ {"label": label, "iso": datetime.combine(d, t, _DEMO_TZ).isoformat()} for label, d, t in candidates ]@tooldef schedule_meeting(topic: str, attendee: Optional[str] = None) -> str: """Ask the user to pick a time slot for a call, via an in-chat picker. Args: topic: Short human-readable description of the call's purpose. attendee: Who the call is with (optional). Returns: Human-readable result string describing the chosen slot or indicating the user cancelled. """ # `interrupt()` pauses the LangGraph run and forwards a structured # payload to the client. The frontend v2 `useInterrupt` hook renders # the picker inline in the chat, then calls `resolve(...)` with the # user's selection — that value comes back here as `response`. response: Any = interrupt( { "topic": topic, "attendee": attendee, "slots": _candidate_slots(), } ) if isinstance(response, dict): if response.get("cancelled"): return f"User cancelled. Meeting NOT scheduled: {topic}" chosen_label = response.get("chosen_label") or response.get("chosen_time") if chosen_label: return f"Meeting scheduled for {chosen_label}: {topic}" return f"User did not pick a time. Meeting NOT scheduled: {topic}"model = ChatOpenAI(model="gpt-5.4")graph = create_agent( model=model, tools=[schedule_meeting], middleware=[CopilotKitMiddleware()], system_prompt=SYSTEM_PROMPT,)Going headless#
Both patterns above ship with a render prop — CopilotKit handles the
"when to show the picker" logic for you. If you want to drive
interrupt resolution from a custom UI that lives anywhere in the tree
(not necessarily inside a chat), see the
headless interrupts guide — it shows
how to compose useAgent, agent.subscribe, and copilotkit.runAgent
to build your own useInterrupt equivalent.
